Research team, data, and aims. This revision, the last one permitted for this application, represents a multi- disciplinary effort that uses specimens and data from 3 large population-based studies of myocardial infarction (Ml), sudden death, and stroke in patients with treated hypertension. The major aims are: (1) to identify new regions that are candidates for drug-gene interactions for each of 4 major anti-hypertensive drug classes on these cardiovascular outcomes;and (2) to replicate the findings to assess validity. Methods. The proposed study has five major steps. Together, Steps 1 to 3 represent a single-stage case-only design to identify genomic regions. In Step 1, a whole-genome case-only study will assay 317,000 SNPs in 1400 cases to identify 150 "interesting" genomic regions for each of the 4 major drug classes (600 for all 4 drug groups [diuretics, beta-blockers, ACE inhibitors, calcium antagonists]). In Step 2, these 600 SNPs will be genotyped in 1400 controls to verify the case-only assumption of no drug-gene association in the population. In Step 3, for each high-signal region, 4 nearby SNPs will be selected and genotyped in the 1400 cases to identify regions where the signal becomes brighter and thus provide empiric support for the selection of the top 60 regions (15 per drug class). They will be selected for further study on the basis of extreme p-values, population attributable fractions, and bioinformatics. In Step 4, HapMap data, resequencing of 20 genes, and other resources will be used to select ethinic-specific tag-SNPs for each of the 60 regions to fully characterize the genetic variation. In Step 5, these tag-SNPs will be genotyped in 3 populations for replication. In the internal replication study, a fresh sample of 600 cases and 1200 controls from the Group Health population will be used to evaluate the drug-gene interactions. In the external replication study, the same genotypes will be assayed in participants with treated hypertension-2700 from the Cardiovascular Health Study and 2000 from the Jackson Heart Study. The revised single-stage case- only design replaces the previous two-stage case-only / case-control study. The new approach is at once more powerful and less expensive than the previous one. This genome-wide association study, which serves as a complement to on-going candidate-gene approaches, has excellent power to detect and replicate modest-sized interactions of antihypertensive drugs with common genetic variants on CV events.